Since 1997, highly pathogenic avian influenza viruses of the H5N1 subtype have infected humans with high case fatality rates, although no sustained human-to-human transmission has yet been reported. Currently, the molecular features and mechanisms that would result in human-to-human transmission of H5N1 viruses are not fully understood. Indeed, several attempts in the past to select transmissible H5 viruses (which typically do not transmit among mammals) were not successful, suggesting that influenza virus transmissibility is determined by several currently unknown factors. Recently, we screened H5 virus libraries possessing random mutations in the globular head region of the hemagglutinin (HA) protein and identified mutant H5 HAs that acquired the ability to bind to human-type receptors. These mutant H5 HAs did not support virus transmission among ferrets (an established influenza virus transmission model) via respiratory droplets, but acquired this ability after two passages of the virus in these animals, which resulted in the selection of additional mutations in HA. This marks the first conversion of an H5 virus that does not transmit among ferrets into one with efficient respiratory droplet transmission. Based on this finding, we propose to decipher the determinants of H5N1 virus transmission in mammals. In Aim 1, we plan To Identify the Mechanisms That Control H5N1 Virus Transmissibility in Mammals. To gain a better understanding of the mutations in HA that result in transmissible viruses, we will select transmissible viruses based on H5 HA proteins derived from different subclades that have caused human infections. Our recent study suggested that HA stability may contribute to virus transmissibility. To test this concept, we also plan to identify mutations in HA that increase HA stability and then test these mutations for their significance in virus transmissibility. The HA proteins of all ferret-transmissible H5N1 viruses will then be characterized for their receptor-binding specificity, their structural consequences, their effects in other genetic backgrounds, and their pathogenicity in mice and ferrets. Mutations in HA that allow avian influenza viruses to bind to human-type receptors are most likely a prerequisite for transmission among mammals; however, findings by us and others indicate that human-type receptor binding is not sufficient for respiratory droplet transmission among ferrets, and that other viral genes also contribute to transmissibility. In Aim 2 (To Characterize the Contribution of Viral Genes Other than HA to H5N1 Virus Transmissibility, we plan to passage non-transmissible viruses of different genetic backgrounds in ferrets to select transmissible mutants. Selected mutations will be characterized for their biological effects, using established assays for internalization, intracellular transport replication and transcription, assembly and budding, and interferon antagonism. Collectively, these studies are expected to generate critical information about the molecular determinants and mechanisms of H5N1 virus transmissibility in mammals.